Search

As malaria incidence decreases and more countries move towards elimination, maps of malaria risk in low‐prevalence areas are increasingly needed. For low‐burden areas, disaggregation regression models have been developed to estimate risk at high spatial resolution from routine surveillance reports aggregated by administrative unit polygons. However, in areas with both routine surveillance data and prevalence surveys, models that make use of the spatial information from prevalence point‐surveys might make more accurate predictions. Using case studies in Indonesia, Senegal and Madagascar, we compare the out‐of‐sample mean absolute error for two methods for incorporating point‐level, spatial information into disaggregation regression models. The first simply fits a binomial‐likelihood, logit‐link, Gaussian random field to prevalence point‐surveys to create a new covariate. The second is a multi‐likelihood model that is fitted jointly to prevalence point‐surveys and polygon incidence data. We find that in most cases there is no difference in mean absolute error between models. In only one case, did the new models perform the best. More generally, our results demonstrate that combining these types of data has the potential to reduce absolute error in estimates of malaria incidence but that simpler baseline models should always be fitted as a benchmark. [ABSTRACT FROM AUTHOR]

Malaria transmission in Madagascar is highly heterogeneous, exhibiting spatial, seasonal and long-term trends. Previous efforts to map malaria risk in Madagascar used prevalence data from Malaria Indicator Surveys. These cross-sectional surveys, conducted during the high transmission season most recently in 2013 and 2016, provide nationally representative prevalence data but cover relatively short time frames. Conversely, monthly case data are collected at health facilities but suffer from biases, including incomplete reporting and low rates of treatment seeking. We combined survey and case data to make monthly maps of prevalence between 2013 and 2016. Health facility catchment populations were estimated to produce incidence rates from the case data. Smoothed incidence surfaces, environmental and socioeconomic covariates, and survey data informed a Bayesian prevalence model, in which a flexible incidence-to-prevalence relationship was learned. Modelled spatial trends were consistent over time, with highest prevalence in the coastal regions and low prevalence in the highlands and desert south. Prevalence was lowest in 2014 and peaked in 2015 and seasonality was widely observed, including in some lower transmission regions. These trends highlight the utility of monthly prevalence estimates over the four year period. By combining survey and case data using this two-step modelling approach, we were able to take advantage of the relative strengths of each metric while accounting for potential bias in the case data. Similar modelling approaches combining large datasets of different malaria metrics may be applicable across sub-Saharan Africa.